One of the goals of the human genome initiative is the characterization of the mouse as a model system. We have been involved in the development of a high density genetic map of the mouse by analysis of several multilocus crosses derived from interspecies or intersubspecies matings. DNAs from these mice have been typed for over 1200 loci, about half of which have also been genetically mapped by other investigators. This permits us to map newly defined genes to specific positions in the genetic map, and to integrate our data into composite maps of each chromosome. These studies have, most recently, resulted in the genetic mapping of genes encoding chemokine receptors, heat shock proteins, homeobox genes, mucin, and defensin. Specific map locations can be useful for several reasons. First, such information contributes to the comparative analysis of homologous genes in man and mouse. For example, we mapped the mouse homolog of a human X-linked homeobox gene to chromosome 3. This represents another example in which pseudoautosomal X-linked genes in one species are autosomal in another and represents the first example of such genes on Chr 3. Second, proximity to a known developmental mutation can identify such a gene as a potential candidate for the abnormal phenotype. We recently mapped a gene involved in bone formation to a site on chromosome 2 near 3 developmental mutations which result in limb deformities. We evaluated the potential role of this candidate gene in separate crosses, but found it to be unrelated to all three mutations. We have recently begun using physical mapping techniques. We are using pulsed field gels to determine gene order and distance for genes not separable by recombination. We are also using a radiation hybrid DNA panel to develop an integrated physical and genetic map of Chr 5.